6. ANALYTICAL METHODS · PDF file6. ANALYTICAL METHODS ... (Durrell and Sauer 1990) to low...
Transcript of 6. ANALYTICAL METHODS · PDF file6. ANALYTICAL METHODS ... (Durrell and Sauer 1990) to low...
MIREX AND CHLORDECONE 237
6. ANALYTICAL METHODS
The purpose of this chapter is to describe the analytical methods that are available for detecting, and/or
measuring, and/or monitoring mirex and chlordecone, their metabolites, and other biomarkers of
exposure and effect to mirex and chlordecone. The intent is not to provide an exhaustive list of
analytical methods. Rather, the intention is to identify well-established methods that are used as the
standard methods of analysis. Many of the analytical methods used for environmental samples are the
methods approved by federal agencies and organizations such as EPA and the National Institute for
Occupational Safety and Health (NIOSH). Other methods presented in this chapter are those that are
approved by groups such as the Association of Official Analytical Chemists (AOAC) and the
American Public Health Association (APHA). Additionally, analytical methods are included that
modify previously used methods to obtain lower detection limits, and/or to improve accuracy and
precision.
6.1 BIOLOGICAL SAMPLES
The most commonly used methods for measuring mirex in blood, tissues (including adipose tissue),
milk, and feces are gas chromatography (GC) or capillary GC combined with electron capture
detection (ECD) or mass spectrometry (MS). Tables 6-l and 6-2 summarize the applicable analytical
methods for determining mirex and chlordecone, respectively, in biological fluids and tissues. Sample
preparation for biological matrices involves solvent extraction followed by clean-up steps. Biological
samples are often contaminated with other compounds such as polychlorinated biphenyls (PCBs);
therefore, additional clean-up steps and/or confirmation techniques are employed to assure reliable
results.
Mirex can be extracted from blood using hexane, acetone-hexane, hexane-ethyl ether, or petroleum
ether and acetone (Bristol et al. 1982; Caille et al. 1987; Korver et al. 1991; Stahr et al. 1980;
Waliszewski and Szymczynski 1991). Blood samples are often contaminated with other compounds
such as PCBs. The use of adsorption chromatography as a clean-up step is effective in achieving
separation of PCBs from mirex in blood (Korver et al. 1991). Other clean-up methods for blood and
tissue samples include concentrated sulfuric acid wash (Waliszewski and Szymczynski 1991), and
Florisil column clean-up (Mes 1992). For measuring mirex in blood, sensitivity of GC/ECD is in the
sub-parts per billion (ppb) range (Korver et al. 1991). Recovery of mirex from blood is generally
MIREX AND CHLORDECONE 2426. ANALYTICAL METHODS
good (≥70%) (Caille et al. 1987; Korver et al. 1991; Stahr et al. 1980; Waliszewski and Szymczynski
1991). Precision is generally very good for blood samples (≤10% relative standard of deviation
[RSD]) (Korver et al. 1991; Stahr et al. 1980). The low RSDs indicate good repeatability of the
procedures (Waliszewski and Szymczynski 1991). Sample storage may adversely affect recovery
(Bristol et al. 1982) and precision (Bristol et al. 1982; Stahr et al. 1980). Confirmation of mirex in
blood can be accomplished by using GC/MS (Korver et al. 1991; Mes 1992).
Mirex can be extracted from tissues using hexane, hexane-acetone, hexane-ethyl ether, or petroleum
ether (Caille et al. 1987; EPA 1980e; Head and Burse 1987; Kutz et al. 1985; LeBel and Williams
1986). Clean-up methods include liquid-liquid partitioning (adipose tissue) (Kutz et al. 1985), gel
permeation chromatography (GPC) (adipose tissue) (LeBel and Williams 1986; Macleod et al. 1982),
and Florisil column clean-up (liver and adipose tissue) (EPA 1980e; Kutz et al. 1985; Mes 1992;
Macleod et al. 1982; Stein and Pittman 1979). For measuring mirex in tissues, sensitivity of GC/ECD
is in the sub-ppm to sub-ppb range (Kutz et al. 1985; LeBel and Williams 1986; Mes 1992; Stein and
Pittman 1979). Recovery of mirex from tissues is generally good (≥70%) (Caille et al. 1987; EPA
1980d; LeBel and Williams 1986; Macleod et al. 1982), as is precision (<20% RSD) (EPA 1980d;
Caille et al. 1987; LeBel and Williams 1986). Confirmation of mirex in adipose tissue can be
accomplished using GC/MS (Kutz et al. 1985; LeBel and Williams 1986; Mes 1992). Photomirex has
been measured in adipose tissue by GC/MS (LeBel and Williams 1986).
Capillary GC/ECD, dual column capillary GC/ECD, and capillary GC/MS have been used for
quantitation of mirex in milk with sensitivity in the low to sub-ppb range (Bush et al. 1983b; Mes et
al. 1986; Mussalo-Rauhamaa et al. 1993; Rahman et al. 1993). Recovery data for milk are generally
very good (>70%) (Mes et al. 1993; Mussalo-Rauhamaa et al. 1993; Rahman et al. 1993), but
precision data were not reported.
Mirex can be extracted from feces with hexane-acetonitrile and the extract cleaned up on
alumina/Florisil columns, then analyzed using GCYECD. Sensitivity, precision, and accuracy data for
feces were not reported (Gibson et al. 1972).
The most commonly used method for measuring chlordecone in blood is GC combined with ECD
(Blanke et al. 1977; Caille et al. 1987; Caplan et al. 1979; Waliszewski and Szymczynski 1991).
MIREX AND CHLORDECONE 243
6. ANALYTICAL METHODS
Sample preparation involves an extraction procedure. Chlordecone is unique among the chlorinated
pesticides since it has a ketone functional group that readily forms a hydrate in the presence of water
(Caplan et al. 1979). This hydrate formation permits selective extraction of chlordecone from all other
chlorinated pesticides (Caplan et al. 1979). Although recoveries for the selective extraction procedure
were low (<50%) because multiple extractions were not performed, sensitivity was maintained and
precision was good (<7% RSD) (Caplan et al. 1979). Another preparation step that allowed better
recovery (>80%) of chlordecone from blood involved extraction with petroleum ether and acetone
followed by a sulfuric acid clean-up step (Waliszewski and Szymczynski 1991). Results of this
method were reproducible, with precision being <7% RSD (Waliszewski and Szymczynski 1991).
Sensitivity was not reported for this method (Waliszewski and Szymczynski 1991). Extraction of
plasma and tissues with hexane-acetone gave low-to-adequate recoveries (58.5-87.4%), but again,
reproducibility was good, with precision being <6% RSD (Caille et al. 1987). Method detection limits
for measuring chlordecone in blood samples are in the low ppb range (Caille et al. 1987; Caplan et al.
1979). Confirmation techniques for chlordecone include GC/MS and GC with microcoulometric
detection (Blanke et al. 1977), and for chlordecone and its breakdown products, GC/chemical
ionization (CI) MS (Harless et al. 1978).
Chlordecone can be extracted from tissues with hexane-acetone, then analyzed by GC/ECD.
Sensitivity is 1 ppm, and recoveries of 73.2% (liver) and 58.5% (kidney) were reported (Caille et al.
1987). No methods for measuring chlordecone in human milk were located.
GC/ECD is the most commonly used method to measure chlordecone in urine and saliva, and
chlordecone and its metabolites (chlordecone alcohol and the glucuronide conjugates) in feces and bile
(Blanke et al. 1977; Fariss et al. 1980). For the liquid samples, using acetone in hexane to extract
chlordecone from acidified samples gave good recoveries (95%) and required no clean-up step (Blanke
et al. 1977). Stool and bile samples required a clean-up procedure prior to analysis. Sensitivity was
5 ppb. For the bile samples, precision was adequate (<20% RSD) (Blanke et al. 1977). No other data
were reported. Chlordecone and its metabolites (chlordecone alcohol and the glucuronide conjugates)
were detected by GC/ECD in feces and bile (Blanke et al. 1978; Fariss et al. 1980). Chlordecone
alcohol was isolated from feces (Wilson and Zehr 1979).
MIREX AND CHLORDECONE 244
6. ANALYTICAL METHODS
6.2 ENVIRONMENTAL SAMPLES
Methods exist for determining mirex and chlordecone in air (ambient and occupational), water,
sediment and soil, biota and fish, and foods. Most involve separation by GC with detection by ECD
or MS. Tables 6-3 and 6-4 summarize some of the applicable analytical methods used for determining
mirex and chlordecone, respectively, in environmental samples.
The most commonly used methods for measuring mirex or its degradation products in air are packed
column or capillary GC/ECD. Air samples are collected using polyurethane foam (PUF), then the
PUF plugs are Soxhlet-extracted (Durrell and Sauer 1990; ASTM 1991; Lewis et al. 1977). For air
samples, sensitivity of GC/ECD is in the sub-ppb range (Durrell and Sauer 1990). Recovery is
excellent (>98%), although precision was not reported (Lewis et al. 1977). Confirmation of mirex
may be accomplished using GC/MS (ASTM 1991) or dual capillary column GC/dual detector (Durrell
and Sauer 1990).
Mirex has been measured in water samples using GC and capillary GC coupled with ECD or MS
detection (Driscoll et al. 1991; Durrell and Sauer 1990; Hargesheimer 1984; Sandhu et al. 1978).
Samples are extracted with dichloromethane (Hargesheimer 1984) or hexane (Driscoll et al. 1991;
Sandhu et al. 1979). Clean-up methodologies which have been applied to water samples are chromic
acid treatment (Driscoll et al. 1991) and Florisil column fractionation (Sandhu et al. 1978). For water
samples, sensitivity is in the low ppb (Durrell and Sauer 1990) to low parts per trillion (ppt) range
(Hargesheimer 1984; Sandhu et al. 1978). Precision is acceptable (<20% RSD) (Driscoll et al. 1991;
Dun-e11 and Sauer 1990; Sandhu et al. 1978). The sensitivity of GC/MS analysis is in the sub-ppb
range (Hargesheimer 1984); recovery and precision data were not reported (Hargesheimer 1984). A
chromic acid digestion extraction technique was compared to conventional solvent extraction for
recovery of mirex and photomirex from river water samples (Driscoll et al. 1991). The digestion
technique was more efficient than conventional solvent extraction, with better recoveries and superior
precision (Driscoll et al. 1991). The better precision obtained with sample digestion may be due to
lack of emulsions, which allowed better phase separation and, therefore, more reproducible recoveries
(Driscoll et al. 1991). Sensitivity data were not reported. Confirmation can be accomplished using
dual capillary GC/dual detector system (ECD and electrolytic conductivity detector, ELCD) (Durrell
and Sauer 1990).
MIREX AND CHLORDECONE 2516. ANALYTICAL METHODS
Mirex and photomirex have been measured in soil and sediment samples using GC and capillary
GC/ECD. Soil and sediment samples are usually solvent extracted, then cleaned up using Florisil
columns and GPC. Recovery of mirex and photomirex from sediment samples is generally excellent
(>90%) (Chau and Babjak 1979; Norstrom et al. 1980a; Onuska et al. 1980; Sergeant et al. 1993) with
very good precision (<20 %RSD) (Norstrom et al. 1980a; Onuska et al. 1980). Sensitivity is in the
low ppb to low ppt range (Norstrom et al. 1980a; Sergeant et al. 1993).
Mirex and its degradation products have been measured in biota using GC/ECD, capillary GC/ECD
and capillary GC/MS techniques (Bush and Barnard 1982; Hellou et al. 1993; Norstrom et al. 1980a;
Onuska et al. 1980; Quintanilla-Lopez et al. 1992). Samples are homogenized and most commonly
extracted with solvent shake-out (Hellou et al. 1993; Norstrom et al. 1980a) or Soxhlet extraction
(Quintanilla-Lopez et al. 1992). The clean-up techniques that are most commonly used are Florisil
columns (Bush and Barnard 1982: Hellou et al. 1993; Norstrom et al. 1980a) and GPC (Hellou et al.
1993; Norstrom et al. 1980a). An additional nitration step has been used to separate mirex and
photomirex from PCBs (Norstrom et al. 1980a). Sensitivity of GC/ECD analysis is in the low to subppb
range. Recoveries are excellent (>90%), and precision is good (<20% RSD). Mirex and its
degradation products have been measured in gull eggs using GC/ECD with capillary GC/MS
confirmation (Norstrom et al. 1980b).
Packed and capillary GC/ECD or GC/MS have been used to measure mirex in foods, including fruits,
vegetables, and fatty foods (Bong 1977; de la Riva and Anadon 1991; Di Muccio et al. 1991; Krause
1973; Liao et al. 1991; Manes et al. 1993; Stan 1989; Trotter and Dickerson 1993). Food samples are
most commonly homogenized and extracted with solvent, then cleaned up using GPC (Stan 1989;
Trotter and Dickerson 1993), Florisil columns (de la Riva and Anadon 1991; Krause 1973), or SPE
columns (Di Muccio et al. 1993; Manes et al. 1993). Sensitivity is in the low to sub-ppb range for
both GC/ECD and GC/MS techniques (de la Riva and Anadon 1991; Liao et al. 1991; Manes et al.
1993; Trotter and Dickerson 1993). Good to excellent recovery (>85% to >90%) and good precision
(<20% RSD) were obtained for most methods (Bong 1977; Di Muccio et al. 1991; Trotter and
Dickerson 1993). Confirmation was accomplished using a different capillary column (Manes et al.
1993; Trotter and Dickerson 1993). GC/ECD has been used to measure mirex in fatty foods with
excellent recovery and good precision; however, the method is not suitable when PCBs are present
(Ault and Spurgeon 1984).
MIREX AND CHLORDECONE 252
6. ANALYTICAL METHODS
The major analytical problem in the measurement of mirex and photomirex in environmental samples
is co-elution with interferents. Confirmation techniques have been developed to assure reliable results.
A dual-column, dual-detector GC analysis has been used to prevent false-positive identifications due to
interfering compounds and to avoid misidentification (Durrell and Sauer 1990). The two detectors
used were ECD and ELCD. MS techniques have been used to assure correct identification
(Hargesheimer 1984; Hellou et al. 1993; Liao et al. 1993; Onuska et al. 1980; Stan 1989) and also to
confirm GC/ECD measurements (Sergeant et al. 1993). Chemical procedures have been used as well.
Perchlorination (Hallett et al. 1978) and nitration (Norstrom et al. 1980a, 1980b) have been used to
convert co-eluting PCBs to compounds easily separable from mirex.
The most commonly used methods for measuring chlordecone and its degradation products in air,
water, soil, sediment, fish, shellfish, and animal fat are similar to those used for mirex (i.e., GC/ECD
techniques and confirmation by GC/MS). Because of the polar nature of chlordecone, the removal of
chlordecone from the different types of environmental samples was accomplished using extraction with
polar solvents (Moseman et al. 1977). The clean-up steps generally used for the environmental
samples include Florisil column chromatography and GPC.
Air samples are collected using filters, or filters and impingers, and extracted with benzene and
methanol (Hodgson et al. 1978; NIOSH 1984). Sensitivity is in the low ppb range for GC/ECD.
Recovery is very good (>85%); precision is acceptable (<25% RSD) (Hodgson et al. 1978; NIOSH
1984). Confirmation of the identity of chlordecone in air was accomplished using both GC/MS and
GC/ELCD (Hodgson et al. 1978).
Water samples are usually solvent extracted and may be analyzed directly by GC/MS (Spingam et al.
1982). Sensitivity is in the low ppb range, but recovery is low (7-11%) and precision is poor (48%
RSD). Extracts may be cleaned up on Florisil columns and analyzed by GC/ECD (Garman et al.
1987; Moseman et al. 1977; Saleh and Lee 1978). Recoveries were very good (>90%) with sensitivity
of GC/ECD being in the low to sub-ppt range (Harris et al. 1980; Moseman et al. 1977; Saleh and Lee
1978); precision data were not reported. Detection limits were lowered to sub-ppt levels by passing
large volumes of water through XAD-2 resin, then extracting the resin (Harris et al. 1980). Recovery
was very good (91%) as was precision (4% RSD).
MIREX AND CHLORDECONE 253
6. ANALYTICAL METHODS
Sediment and soil samples are homogenized and extracted. Clean-up procedures are required prior to
analysis by GC/ECD or GC/MS techniques (Lopez-Avila et al. 1992; Moseman et al. 1977; Saleh and
Lee 1978; Tieman et al. 1990). For sediment, soil, and sludge, recoveries were good (>85%) with
sensitivity in the low ppb range (Moseman et al. 1977; Saleh and Lee 1978). Precision is good (<6%
RSD) (Saleh and Lee 1978). Analytical difficulties (unacceptable recovery; not detectable using
second capillary GC column) were reported (Lopez-Avila et al. 1992; Tieman et al. 1990).
Fish samples are extracted and cleaned up using liquid-liquid partitioning or Florisil columns prior to
analysis by GC/ECD (Carver and Griffith; Hodgson et al. 1978; Mady et al. 1979; Moseman et al.
1977). Recoveries are good for chlordecone (> 80%) (Carver and Griffith 1979; Hodgson et al. 1978;
Mady et al. 1979) and the monohydro and dihydro degradation products (Carver and Griffith 1979).
Precision is good (Carver and Griffith 1979; Mady et al. 1979) and sensitivity is in the low ppb range
(Carver and Griffith 1979; Mady et al. 1979).
Few methods for measuring chlordecone in foods are available. Lower recoveries (58-81%) were
obtained with GC/ECD for beef, pork, and poultry fat samples using GPC clean-up before analysis
(Goodspeed and Chestnut 1991). Precision varied greatly (7.1-47.7% RSD) because of the lower
recoveries; sensitivity was not reported.
6.3 ADEQUACY OF THE DATABASE
Section 104(i)(5) of CERCLA, as amended, directs the Administrator of ATSDR (in consultation with
the Administrator of EPA and agencies and programs of the Public Health Service) to assess whether
adequate information on the health effects of mirex and chlordecone is available. Where adequate
information is not available, ATSDR, in conjunction with the NTP, is required to assure the initiation
of a program of research designed to determine the health effects (and techniques for developing
methods to determine such health effects) of mirex and chlordecone.
The following categories of possible data needs have been identified by a joint team of scientists from
ATSDR, NTP, and EPA. They are defined as substance-specific informational needs that if met would
reduce the uncertainties of human health assessment. This definition should not be interpreted to mean
that all data needs discussed in this section must be filled. In the future, the identified data needs will
be evaluated and prioritized, and a substance-specific research agenda will be proposed.
MIREX AND CHLORDECONE 254
6. ANALYTICAL METHODS
6.3.1 Identification of Data Needs
Methods for Determining Biomarkers of Exposure and Effect. There are reliable methods
for detecting, quantifying, and identifying mirex and chlordecone in biological samples. These include
packed column and capillary GC/ECD and packed column and capillary GC/MS. These methods are
sensitive enough to measure background levels in the population and levels at which biological effects
occur. These methods are accurate and reliable for measuring mirex in blood (Korver et al. 1991; Mes
1992) and chlordecone in blood (Caille et al. 1987). Sensitivity for these methods is in the low to
sub-ppb range. Sensitive (low to sub-ppb range) and accurate methods are available to measure mirex
in tissues (Caille et al. 1987; LeBel and Williams 1986; Mes 1992). Improved recovery data and
greater sensitivity for measuring chlordecone in tissues are needed (Caille et al. 1987). For milk,
fecal, bile, urine, and saliva samples, sensitivity, recovery, and precision data are needed to more fully
evaluate the reliability of these methods as predictors of environmental exposure to both mirex and
chlordecone (Blanke et al. 1977; Bush et al. 1983b; Gibson et al. 1972).
Biochemical indicators of renal dysfunction (increased urinary protein and/or histopathological changes
of the kidneys) have been associated with exposure to both mirex (NTP 1990) and chlordecone
(Larson et al. 1979b). Microsomal enzyme induction as shown by changes in urinary D-glucaric acid
has also been associated with exposure to both mirex and chlordecone (Guzelian 1985; Morgan and
Roan 1974). Although these changes are not specific for mirex or chlordecone, these parameters may
provide information about renal damage and hepatic effects in exposed populations. Tremorgrams
have been used to assess tremors associated with chlordecone exposure in humans (Taylor et al. 1978).
An infrared reflection technique and oculography have been used to assess the oculomotor disturbances
caused by chlordecone (Taylor et al. 1978). Standard tests for memory and intelligence can be used to
determine the presence of encephalopathy, but in the absence of baseline pre-exposure levels for
individuals, subtle changes may be difficult to detect. Decreased sperm count has been observed
following exposure to mirex or chlordecone (Chu et al. 1981a; Yarborough et al. 1981). The existing
analytical methods that are discussed for exposure can reliably measure mirex or chlordecone in blood,
urine, and tissues at the levels at which these effects occur.
MIREX AND CHLORDECONE 255
6. ANALYTICAL METHODS
Methods for Determining Parent Compounds and Degradation Products inEnvironmental Media. Reliable methods for detecting mirex and chlordecone in environmental
media include GC/ECD, capillary GC/ECD and capillary GC/MS. In general, the methods are
sensitive and accurate enough to measure background levels of mirex and chlordecone in the
environment and levels at which health effects occur. Methods of adequate sensitivity (low ppb to
sub-ppb), accuracy, and specificity are available for determining levels of mirex in air (Dun-e11 and
Sauer 1990; Hoff et al. 1992; Lewis et al. 1977), water (Durrell and Sauer 1990; Hargesheimer 1984;
Sandhu et al. 1978), and soils and sediment (Norstrom et al. 1980a; Sergeant et al. 1993; Seidel and
Lindner 1993). Sensitive, accurate methods are also available for measuring chlordecone in air
(NIOSH 1984), water (Garman et al. 1987; Harris et al. 1980; Saleh and Lee 1978: Spingarn et al.
1982), and soil and sediment (Moseman et al. 1977; Saleh and Lee 1978). Methods for measuring
mirex and chlordecone in aquatic species and food are reliable and accurate and provide detection
limits in the low ppm to ppb range. These include methods for determining mirex in fish and other
aquatic species (Bush and Barnard 1982; Hellou et al. 1993; Norstrom et al. 1980a; Quintanilla-Lopez
et al. 1992; Rahman et al. 1993) and food (Ault and Spurgeon 1984; Liao et al. 1991; Manes et al.
1993; Trotter and Dickerson 1993). Similarly, there are acceptable methods for determining
chlordecone in fish and other aquatic species (Carver and Griffith 1979; Mady et al. 1979) and food
(Goodspeed and Chestnut 1991; Posyniak and Stec 1980). More information on the precision of these
methods for measuring mirex and chlordecone in water and improved sensitivity, recovery, and
precision data in foodstuffs are needed to better assess the risk of exposure for these media. Research
investigating the relationship between levels of mirex and chlordecone measured in air, water, soil, and
food and observed health effects could increase our confidence in existing methods and/or indicate
where improvements are needed. No data were located regarding measurement of mirex in soil
samples.
6.3.2 Ongoing Studies
Research is being conducted at the State University of New York at Albany, sponsored by the
National Institute of Environmental Health Sciences, to improve chemical analysis of environmental
media for PCBs and selected pesticides, including mirex. No other studies involving rnirex or
chlordecone were located in the FEDRIP database.